NATURE’S GIFT TO SCIENCE Nobel Lecture, December 8, 2002 by SYDNEY BRENNER Salk Institute, 100010 N. Torrey Pines Road, La Jolla, California, USA, and King’s College, Cambridge, England. The title of my lecture is “Nature’s gift to Science.” It is not a lecture about one scientific journal paying respects to another, but about how the great di- versity of the living world can both inspire and serve innovation in biological research. Current ideas of the uses of Model Organisms spring from the ex- emplars of the past and choosing the right organism for one’s research is as important as finding the right problems to work on. In all of my research these two decisions have been closely intertwined. Without doubt the fourth winner of the Nobel prize this year is Caenohabditis elegans; it deserves all of the honour but, of course, it will not be able to share the monetary award. I intend to tell you a little about the early work on the nematode to put it into an intellectual perspective. It bridges, both in time and concept, the biol- ogy we practice today and the biology that was initiated some fifty years ago with the revolutionary discovery of the double-helical structure of DNA by Watson and Crick. My colleagues who follow will tell you more about the worm and also recount their incisive research on the cell lineage and on the genetic control of all death. To begin with, I can do no better than to quote from the paper I published in 1974 (1). The paper was unhesitatingly entitled: “The genetics of Caenor- habditis elegans” and the opening sentence reads: “How genes might specify the complex structures found in higher organisms is a major unsolved prob- lem of biology.” This is still true today. The paper outlined how a genetic ap- proach coupled with detailed studies at the cellular level might be a way of studying this important question. It introduced C. elegans as the organism of choice for this work. This choice had a long history. Twenty years earlier, we had posed a differ- ent question. Then, the central problem in biology was how the one-dimen- sional sequence of nucleotides in DNA specified the one-dimensional se- quence of amino acids in proteins. Today, any student would give this question a very simple answer. ‘All you have to do is to find a gene and have it sequenced and then make some protein using the gene and get someone to determine its amino acid sequence.’ In those early days, the techniques for determining amino acid sequences of proteins were primitive and needed large amounts of proteins which had to be purified first. There were no 274 methods to isolate genes and no techniques for the chemical determination of their sequences. Our analysis of genes was limited to genetics. Indeed, the only way we could assert that there was a gene in an organism was by finding a mutant allele for it. Like Mendel, we could not say that there was a gene for the character of tallness until dwarf mutants were discovered suffering from a heritable lack of tallness. Genetic analysis of linkage used recombination to analyse the structure of chromosomes, to determine the locations of genes and their linear order along a chromosome. But in order to probe the struc- ture of the gene something special was needed. That was provided by bacte- riophages, viruses with tiny genomes which grew on bacteria and which showed recombination. In 1954, Seymour Benzer developed a system using the rII gene of bacteriophage T4, in which powerful selection could be ex- erted for the r+ phenotype. Forward mutants could be easily picked and large numbers of phage crosses could be readily performed. The selection method allowed recombinants to be measured at very high resolution, limited only by the rates of reverse mutations. These experiments revealed that the gene con- tained hundreds of sites at which mutation could occur and that the scale of separation was of the order of the distances between adjacent base pairs on the DNA structure. The experiments not only exploded the classical idea of the gene as an indivisible unit of function, mutation and recombination, but the fine structure map could now be viewed as a picture of the nucleotide se- quence of the DNA. It therefore provided an approach to studying how the sequence of bases in DNA might correspond to the sequence of amino acids in proteins by using genetics for the first and chemistry for the latter. This programme of molecular genetics was never quite completed, but the ex- ploitation of the properties of T4 bacteriophages played important roles in studies of mutagenesis, the general nature of the genetic code and genetic suppression, and in the demonstration of the existence of messenger RNA. For our purposes today we need to reflect on the properties that made phages that ideal ‘model organisms’ for this phase of research in molecular biology. They were easy to grow and maintain in a laboratory, large numbers could be readily generated, and many experiments could be conducted in parallel. The final readout was the presence or absence of lysis of bacteria and this assay could be applied to single particles each of which left a plaque – an area of lysis in a lawn of bacteria. For mapping purposes, scoring was digital – yes or no – and single particles could be easily counted to obtain accurate fre- quencies. After the basic principles of information transfer from genes to proteins had been established with the identification of messenger RNA, the discovery of the mechanism of protein synthesis and the structure of genetic code, it was natural for some of us to ask whether the lessons learnt in molecular biol- ogy could be applied to the genetics of more complex phenotypes. All ques- tions in genetics involve asking how the phenotype is represented in the genotype or, reflexively, how the genes map onto the phenotype. It was clear that one could not have understood how metabolism or biosynthesis was rep- resented on the genome until one had conceived of the ‘one gene-one en- 275 zyme relationship’, that a gene specified the amino acid sequence of a polypeptide chain, which after folding correctly, was able to perform a spe- cific catalytic function in a metabolic pathway. In the same way, the elaborate structure of the protein coat of a bacteriophage could not be understood without knowing that it was built of many different kinds of subunits and that these were able to self-assembly into the final particle. We have to know how the gene specifies the construction of the entity which carries out the func- tion. The same is true at a higher level of organization. In studying the ge- netics of behaviour, it is difficult, if not impossible, to go directly from the gene to behaviour, because there is no simple mapping that connects the two. In my paper, I put it in this way: “Behaviour is the result of a complex ill-un- derstood set of computations performed by nervous systems and it seems es- sential to decompose the question into two: one concerned with the question of the genetic specification of nervous systems and the other with the way ner- vous systems work to produce behaviour.” Thus, just as the structure and function of protein molecules is the necessary connection between the genes and metabolism, the link between genes and behaviour resides in under- standing the structure of nervous systems and how they are constructed. These are questions of anatomy and embryological development. This set the requirements for the experimental organism as one which was not only suit- able for genetical study in the laboratory but also allowed the structure of the nervous system to be accurately defined. Since a nervous system is essentially a cellular network, we had to be able to observe junctions between cells and their processes and this could only be achieved with the electron microscope which has the necessary resolution. Since the electron microscope provides only a small window because of its high magnification, we needed a small ani- mal which would also need to have a nervous system with a small number of cells. After some searching, my choice finally settled on the small nematode, Caenorhabditis elegans. This was a self-fertilizing hermaphrodite with rare spon- taneous males. The adults are about 1 mm in length and the life cycle is com- 1 pleted in 3 /2 days. The animals live in a two-dimensional world feeding on E. coli on the surface of agar plates. They are easy to grow in bulk, each ani- mal producing about 300 progeny during a cycle. My paper reported the iso- lation of several hundred mutants together their complementation and map- ping to define about one hundred loci. In parallel with the genetic work, I launched a program with Nichol Thompson to determine the complete struc- ture of the worm by serial section electron microscopy (2). This project was completed by John White and Eileen Southgate who joined the group soon af- ter it started and whose work resulted in the determination of the complete structure of the nervous system (and much more besides) in C.elegans (3). The paper I referred to summarized the work on genetics but, in the early seventies, trying to understand the functions of genes specifying the develop- ment of the nervous system in molecular terms seemed impossibly remote. However, going to the molecular level was inevitable, and we very early initi- ated studies of mutants which affected the movement of the worm by disrupting muscle structure function (4,5,6).